Epigenetic Responses of Hare Barley (Hordeum Murinum Subsp. Leporinum) to Climate Change: an Experimental, Trait-Based Approach

Heredity (2021) 126:748–762 https://doi.org/10.1038/s41437-021-00415-y

ARTICLE

Epigenetic responses of hare barley (Hordeum murinum subsp. leporinum) to climate change: an experimental, trait-based approach

1,2,3 2 1 2 Víctor Chano ● Tania Domínguez-Flores ● Maria Dolores Hidalgo-Galvez ● Jesús Rodríguez-Calcerrada ● Ignacio Manuel Pérez-Ramos1

Received: 12 June 2020 / Revised: 29 January 2021 / Accepted: 29 January 2021 / Published online: 19 February 2021 © The Author(s) 2021. This article is published with open access

Abstract The impact of reduced rainfall and increased temperatures forecasted by climate change models on plant communities will depend on the capacity of plant species to acclimate and adapt to new environmental conditions. The acclimation process is mainly driven by epigenetic regulation, including structural and chemical modiﬁcations on the genome that do not affect the nucleotide sequence. In plants, one of the best-known epigenetic mechanisms is cytosine-methylation. We evaluated the impact of 30% reduced rainfall (hereafter “drought” treatment; D), 3 °C increased air temperature (“warming”; W), and the combination of D and W (WD) on the phenotypic and epigenetic variability of Hordeum murinum subsp. leporinum L.,

1234567890();,: 1234567890();,: a grass species of high relevance in Mediterranean agroforestry systems. A full factorial experiment was set up in a savannah-like ecosystem located in southwestern Spain. H. murinum exhibited a large phenotypic plasticity in response to climatic conditions. Plants subjected to warmer conditions (i.e., W and WD treatments) ﬂowered earlier, and those subjected to combined stress (WD) showed a higher investment in leaf area per unit of leaf mass (i.e., higher SLA) and produced heavier seeds. Our results also indicated that both the level and patterns of methylation varied substantially with the climatic treatments, with the combination of D and W inducing a clearly different epigenetic response compared to that promoted by D and W separately. The main conclusion achieved in this work suggests a potential role of epigenetic regulation of gene expression for the maintenance of homoeostasis and functional stability under future climate change scenarios.

Introduction

There is a growing interest in understanding how plants will adapt to the projected changes in climate. Plant species respond to environmental changes by means of multiple Supplementary information The online version contains morphological and physiological adjustments that serve to supplementary material available at https://doi.org/10.1038/s41437- alleviate stress levels and to increase the uptake of limiting 021-00415-y. resources (Nicotra et al. 2010;Freschetetal.2018;Pérez- * Víctor Chano Ramos et al. 2019). For example, plants subjected to [email protected] greater water/nutrient limitations usually exhibit a suite of trait values associated to efficient resource conservation 1 Research Group “Sistemas Forestales Mediterráneos”, Instituto de Recursos Naturales y Agrobiología de Sevilla. Dpto, use (e.g., plants with small-sized and high-density leaves) Biogeoquímica, Ecología Vegetal y Microbiana, Consejo Superior (Chapin et al. 1993;Wrightetal.2004), which increases de Investigaciones Científicas, Av. Reina Mercedes 10, 41012 competitive abilities under resource-limiting conditions. Sevilla, Spain This strategy contrasts with that displayed by plants 2 Research Group “Sistemas Naturales e Historia Forestal”, ETSI inhabiting moist and fertile sites, with opposite trait values Montes, Forestal y del Medio Natural. Dpto, Sistemas y Recursos related to a rapid return on investment. Moreover, under Naturales, Universidad Politécnica de Madrid, Ciudad Universitaria s/n, 28040 Madrid, Spain non-limiting conditions of soil water and nutrient avail- 3 ability, an increase in air temperature could result in Present address: Department of Forest Genetics and Forest Tree fi Breeding, University of Göttingen, Büsgenweg 2, 37077 thinner leaves with higher speci c leaf area that favour Göttingen, Germany plant carbon uptake and growth (Poorter et al. 2009; Epigenetic responses of hare barley (Hordeum murinum subsp. leporinum) to climate change: an. . . 749

Lamaouietal.2018). Plants can also modify their repro- natural populations differing in a limited (one or two) ductive output and phenology under different environ- number of factors. However, in nature, plants are exposed to mental scenarios of temperature and resource limitation a variety of constraints, which constitute a multidimensional (Valencia et al. 2016; Pérez-Ramos et al. 2020). space where many factors act simultaneously and inter- The species’ ability to vary its phenotypic expression actively (Ibáñez and Schupp 2001; Gómez 2004). Since the across environments is mainly driven by genetic variability, effect of one stress factor on plant performance may be which is in turn originated by sexual reproduction and exacerbated or mitigated by another (Mitchell et al. 2015), random mutations (Ewens 2013). In addition, there are the impact of a combination of various stress factors may evidences that epigenetic processes (i.e., structural and differ from the sum of the impacts caused by each factor chemical modifications on the genome that do not affect the applied individually. As a result, combined stress factors nucleotide sequence) can promote fast and reversible phe- may enhance phenotypic variation and plant fitness despite notypic variations in response to environmental changes the presumably higher costs of plasticity (Lampei 2019). (Bossdorf et al. 2008). These epigenetic mechanisms act in Therefore, an accurate prediction of climate change pro- a switch mode, activating/deactivating gene transcription in jections on plant phenotypic variability requires the con- three different ways: (i) DNA methylation by the covalent sideration of potential additive and interactive effects of binding of methyl groups to cytosine nucleotides (5 mC), different abiotic factors. (ii) regulation of DNA accessibility by histone modification The role of methylation or demethylation processes in and (iii) post-transcriptional regulation by non-coding response to stress has not been fully elucidated and seems to RNAs activity such as microRNAs (Bossdorf et al. 2008). be dynamic in time (Liu et al. 2018). For instance, several In plants, DNA methylation has been found to silence alleles were identified and related to a plastic response to transposable elements (TEs) and gene expression (Bartels climate variation in natural populations of Arabidopsis et al. 2018), with promoter methylated genes having lower thaliana, with diverse genome-wide methylation patterns transcription levels (Zhang et al. 2006; Li et al. 2012). associated to seasonality (Shen et al. 2014). Wang et al. Recent studies have argued that epigenetic mechanisms (2011) detected different patterns of genome-wide methy- could play a relevant role in microevolution under chal- lation/demethylation in rice induced by water limitation, lenging environmental scenarios, such as those promoted by with plants exhibiting 70% reversibility to the original climate change (e.g., Kronholm et al. 2017; Jeremias et al. status after drought cessation. Another study also detected 2018; Münzbergová et al. 2019). In fact, although it is drought-induced responses in rice, with hyper- and hypo- broadly assumed that species with high genetic variability methylation being related to susceptibility and tolerance to have higher adaptation potential for a larger variety of water deficit, respectively (Gayacharan 2013). On the environmental conditions than species with low genetic other hand, DNA-methylation processes have been widely diversity (Anderson et al. 2011), recent studies (e.g., Zhang studied in response to heat stress, especially through the et al. 2016) have demonstrated the importance of epigenetic activity of methyltransferases (see review by Liu et al. diversity for environmental adaptation in plant species 2015). All these studies suggest that methylation patterns in with limited genetic diversity. Notwithstanding, there are response to stress conditions depend on the source of stress, still many gaps on the role of epigenetics as a driver of as well as on the genotype, the tissue and the ontogeny this source of phenotypic variability across contrasted (Bonasio et al. 2010; Tan 2010; Wang et al. 2011; Mastan environments. et al. 2012), and that these factors will likely encompass a Several studies have reported the influence of epigenetic wide range of responsive genes. responses to different environmental conditions on some In the present study we analysed the two main sources of developmental processes. For instance, it has been demon- genotypic variability (i.e., genetic and epigenetic) of an strated that phenotypic variations in Viola cazorlensis and annual grass species dominant in Mediterranean savannah- allopolyploid orchids in response to climate are mediated by like ecosystems, Hordeum murinum subsp. leporinum L. both genetic and epigenetic processes (Herrera and Bazaga (hare barley, hereafter), with great relevance in pasture 2010; Paun et al. 2010). Other works have also focused on dynamics due to its fast growth and high palatability for the epigenetic responses to different sources of stress, both livestock (Hulting and Haavisto 2013). This species is biotic, such as herbivory, disease or plant competition for widely distributed in Europe and Middle East and has been nutrients (Conrath et al. 2002; Tani et al. 2005; Verhoeven declared as moderately invasive in California by the et al. 2010; Jaskiewicz et al. 2011), and abiotic, such as California Plant Invasive Council (Supplementary Data salinity, mechanical stress, heat and drought (Herrera et al. Fig. S1). Plants were experimentally subjected to increased 2012; Latzel et al. 2013; Liu et al. 2015; Alsdurf et al. 2016; temperature, decreased rainfall, and the combination of both Banerjee and Roychoudhury 2017). Most of these studies factors. We also examined the phenotypic responses of this have been conducted under controlled conditions or in species to the different climatic treatments by measuring 750 V.íctor Chano et al. eight functional traits related to plant morphology, phenol- warming and drought (WD), plus a control treatment (C) of ogy and reproductive ability. The effort of combining epi- plants subjected to natural conditions. The four treatments genetic analysis under contrasting climatic conditions with were replicated in six plots of 4 × 6 m, with a minimum plant phenotypic characterisation using attributes related distance of 20 m from each other, and fenced with a metallic with the three leading dimensions of ecological variation structure to avoid livestock access (Supplementary Data (i.e., plant economics, light interception, and reproductive Fig. S2). The drought treatment was created by placing 6 ability) (Westoby 1998; Laughlin et al. 2010; de la Riva methacrylate gutters, 0.14-m wide each, inclined 20° over et al. 2016) aims to increase our understanding on the role half of every plot (2.5 × 2.5 × 1.5 m). These rainout shelters of epigenetic regulation as a driver of plant acclimation to reduced the total amount of rainfall reaching the soil surface ongoing climate changes. In fact, global change models by around 33%. To recreate warming conditions, hexagonal predict an increase in temperature and more severe and open-top chambers (OTC) were used (see Marion et al. recurrent drought periods for the next decades in many 1997). They were built with methacrylate sheets without temperate and Mediterranean ecosystems (Lindner et al. UV-Filter to avoid modifying the light spectrum, with 2010). In this context, savannah-like ecosystems in the sloping sides of 40 × 50 × 32 cm, and wavelength trans- Iberian Peninsula (known as dehesa in Spain and montados mission between 280 and 750 nm. Previous studies indi- in Portugal) may be particularly sensitive to increased cated that OTCs increase air temperature by 1–3 °C relative temperatures and decreased rainfall (Olea and San Miguel to the external environment without altering light trans- 2006). These environmental changes could potentially jeo- mission within the chambers (Dabros and Fyles 2010; pardise the weak equilibrium of their plant communities, Aragón-Gastélum et al. 2014). In each plot, two OTCs were which already suffer the inclement conditions of low soil placed under the rainout shelters in order to evaluate the water availability and high temperatures in summer. We impact of temperature increase and rainfall exclusion hypothesise that hare barley will display different epigenetic simultaneously (WD treatment) and two outside (W treat- and functional responses to the different environmental ment). These permanent structures of rainfall exclusion and scenarios, with the combination of increased temperature temperature rise were chosen due to their successful use in and reduced rainfall resulting in a differential response past studies on climate change simulations (e.g., Delgado- compared with the responses resulting from standalone Baquerizo et al. 2013; Maestre et al. 2013). climatic treatments. More specifically, we hypothesise that Soil volumetric water content was quantified using a plants subjected to warmer conditions will exhibit earlier capacitance soil moisture probe (Delta T Devices). In half of reproductive phenology and higher plant height, and plants the experimental plots, one 40-cm long tube was inserted subjected to decreased rainfall will show trait values asso- into the soil per climatic treatment, and measurements of ciated to higher efficient resource conservation (i.e., small- soil humidity were taken periodically from November to sized and high-density leaves). April. Soil surface air temperature was quantified hourly by means of HOBO-type sensors (Alpha Omega Electronics) in half of the experimental plots (three replicates per cli- Material and methods matic treatment). As expected, air temperature was higher in the experimental units located within the OTCs (i.e., W and Study site and experimental conditions WD treatments), whereas soil humidity decreased in the units located beneath the rainout shelters (i.e., D and WD The study was carried out in “La Morra” (X:340933, treatments; see Supplementary Data Fig. S3). Y:4246190), a dehesa located in southwestern Spain (Los Pedroches valley, Córdoba). Climate is Mediterranean, with Plant phenotypic measurements cool wet winters and hot dry summers. Mean annual rainfall is 428.2 mm, and mean annual temperature is 14.6 °C (with In 24 experimental units (i.e., 6 plots × 4 climatic treat- a mean monthly maximum of 34.7 °C in July and a mean ments), four 21 × 21 cm PVC quadrats (divided in turn into monthly minimum of 0.7 °C in January; mean values for the nine squares of 7 × 7 cm) were placed at the peak of max- 2008–2017 period). Vegetation is characterised by a dense imum vegetative growth (i.e., in the mid-spring of 2017 and cover of herbaceous species (≥ 65% relative plant cover; 2018) to determine species abundance and composition. most of them annuals), which coexist with scattered oaks Species frequencies were calculated from the number of (mostly Quercus ilex; ~20%) and shrub species (≤7.5%, squares where each species was present. The relative fre- mostly Crataegus monogyna and Retama sphaerocarpa). quency of hare barley and its temporal evolution from In September 2016, before the beginning of the rainy 2017 to 2018 was further compared to those of the co- season, a factorial experiment was designed with four cli- dominant species Avena barbata, Crepis capillaris, Ero- matic treatments: warming (W, hereafter), drought (D), dium moschetum, Geranium dissectum and Sinapis alba. Epigenetic responses of hare barley (Hordeum murinum subsp. leporinum) to climate change: an. . . 751

These five species were selected because they were among used for DNA extraction using the CTAB method (Doyle the most abundant in the different experimental units and and Doyle 1987). The isolated DNAs were quantified with a their relative frequencies varied widely from the first to the NanoDrop 2000 UV (Thermo scientific). second sampling year. In April 2017, at peak biomass, 10–30 individuals of AFLP analysis hare barley per climatic treatment were randomly selected to measure one whole-plant trait (plant height), one Twenty samples from the C treatment distributed among the reproductive trait (seed mass) and three morphological 6 experimental plots were used for the analysis of genetic above-ground traits: leaf size, specific leaf area (SLA; leaf variation between plots by using Amplified Fragment area per unit of leaf dry mass), and leaf dry matter content Length Polymorphism (AFLP). This tool is based on the (LDMC; dry mass per unit of water-saturated fresh mass). analysis of DNA markers resulting from the fragmentation All these traits were measured at the level of treatment of DNA using EcoRI and MseI restriction enzymes (Vos instead of at the plot level due to the low number of indi- et al. 1995), which cut a specific sequence of nucleotides in viduals present in some plots. Thus, plant height was the DNA. For each individual sample reaction, 500 ng of measured in 30 plants per climatic treatment using a cal- total DNA were digested at 37 °C for 3 h and 100 rpm using liper. Leaf size, SLA and LDMC were quantified in ten 5 U of EcorRI (New England Biolabs) and 5 U of MseI plants per climatic treatment, following the protocols (New England Biolabs) in a final volume of 25 µl. The described by Garnier et al. (2001). Leaf size was quantified DNA fragments were then ligated to double-stranded EcoRI using an image analysis programme (Image Pro-plus 4.5; and MseI adaptors (Supplementary Data Table S1), in a Media Cybernetic Inc., Rockville, MD, USA). Seed mass final volume of 30 µl, using 1 U of T4 DNA ligase (New was quantified by weighing all the seeds contained in one England Biolabs) at 37 °C during 6 h and 100 rpm, and then infrutescence per plant, in at least 10 plants per climatic overnight at 4 °C. In order to subset the number of amplified treatment; seeds were previously oven-dried at 60 °C for fragments, a 20 µl pre-selective PCR reaction was carried 48 h to obtain their dry weight. Additionally, reproductive out using the product of the ligation reaction (for primers phenology at the population level was monitored once a information see Supplementary Data Table S1). PCR con- week over the whole period of flowering of the species ditions were as follows: denaturing at 94 °C for 5 min, 28 (from mid-February to late June 2017). In each census, we cycles of 94 °C for 30 s, 60 °C for 60 s, 72 °C for 60 s, and a counted the number of flowers of hare barley using a semi- final elongation step at 72 °C for 10 min. The product of the quantitative scale (from 0 to 5). We registered the specific first pre-selective PCR was used as template for a sub- dates in which the onset, peak and cessation of flowering sequent 10 µl selective PCR reaction, with the following took place in those experimental units where hare barley conditions: 94 °C for 5 min, 12 cycles of 94 °C for 30 s, was present. The onset of flowering was defined as the date 65–56 °C for 30 s (decreasing 0.7 °C each cycle), 72 °C for when the first flower was observed in at least one indivi- 6 s, 23 cycles of 94 °C for 30 s, 56 °C for 30 s, 72 °C for dual, whereas the cessation of flowering was defined as the 60 s, and a final elongation step of 72 °C for 5 min. This date when the last flower wilted. The difference between selective PCR was performed using 4 primer combinations the onset and the cessation of flowering was used to cal- (Supplementary Data Table S1). Resulting DNA fragments culate the duration of flowering. The peak of flowering was were analysed through the electrophoretic system 4300 defined as the date when the population reached its max- DNA Analyser System (LiCOR Bioscience), along with a imal number of flowers. All these traits were selected 50–1500 bp Size Standard (LiCOR Bioscience). DNA for covering the three dimensions of ecological variation fragments between 100 bp and 500 bp were included into among plants (i.e., plant economics, light interception, and the analysis. According to the presence/absence of bands for reproductive ability) (Westoby 1998; Laughlin et al. 2010; specific positions, referred to as locus, a Boolean datamatrix de la Riva et al. 2016), given their utility in studies of coded by 0 (absence) and 1 (presence) was constructed. abiotic stress and functional ecology (Wright et al. 2004; Pérez-Ramos et al. 2017). MSAP analysis

Sample collection and DNA extraction DNA methylation patterns were identified by Methyl- Sensitive Amplified Polymorphism (MSAP) for the 120 In April 2017, 30 individuals of hare barley per climatic individuals of hare barley collected. This methodology, as treatment (five per plot) were collected, discarding the AFLP, is based on DNA markers from DNA digestion, but inflorescences, and immediately frozen in liquid nitrogen to uses EcoRI/HpaII and EcoRI/MspI couples of restriction avoid DNA methylation due to sampling. In the lab, frozen enzymes in two parallel reactions (Reyna-López et al. 1997). samples were ground, and 100 mg of tissue powder was The main difference between AFLP and MSAP is that with 752 V.íctor Chano et al.

MSAP the isoschizomers HpaII and MspI cut the target epigenetic variation, or Non-Methylated Loci (NML), used depending on the methylation state of the cytosines present as a proxy to assess genetic variation (Watson et al. 2018). in the sequence. As for AFLP, the protocol involved 500 ng Variations among MSL were also calculated with the of total DNA for each reaction, and the same conditions Shannon’s diversity index (S), and epigenetic variations during digestion, adaptors ligation, and PCR. Moreover, a among plots and climatic treatments were also explored by selective PCR was carried out using 12 primer combinations principal coordinate analyses (PCoA) based on Euclidean for each restriction enzyme couple (Supplementary Data distance matrix, implemented in the msap package. Fur- Table S1) and the resulting DNA fragments between 100 bp thermore, an analysis of the molecular variance (AMOVA) and 500 bp were used to construct the Boolean datamatrix. was used for the estimation of variance components and the Phi-statistic (analogue to the F-statistic for binary data) to Detection of global DNA methylation levels reflect genetic and epigenetic diversity (Excoffier et al. 1992). In order to determine those loci showing non- A global 5-methylcytosine analysis (5-mC DNA ELISA randomly distributed methylation patterns between treat- Kit, ZYMO) was used to measure the global DNA methy- ments (h for hemimethylation, i for methylation of inner lation levels in hare barley in response to climatic treat- cytosines, u for non-methylation, and f for uninformative ments. DNA from seven individuals per treatment randomly state, which may be due to full methylation or changes in selected was analysed. The optical density at 405 nm was the nucleotide sequence of the target), a locus-by-locus Chi- determined after 45 min using an Absorbance Microplate squared test was performed using the MSL, following the Reader ELx808TM (Bio-Tek® Instruments, Inc., USA). The reproducible example script for R found in Watson et al. global DNA methylation levels were expressed in percen- (2018). To control the false discovery rate (FDR), a BH tage of DNA as the mean of three technical replicates multitest adjustment was adopted (Benjamini and Hochberg according to the manufacturer instructions. 1995), and those MSL with p-value < 0.001 were considered as significantly differentiated. The relationships Statistical and bioinformatic analysis between significantly differentiated loci were calculated with the Gower’s Coefficient of Similarity (Gower 1971), The effects of the experimental climatic treatments (W, D and the Complex Heatmap package available for Bio- and WD) were tested on the eight phenotypic variables conductor in R (Gu et al. 2016) was used to cluster and related with plant morphology (plant height, leaf size, SLA show the resulting matrix as a heatmap. and LDMC), reproductive output (seed mass) and plant Genetic diversity among the six experimental plots was phenology (flowering onset, peak and duration), as well as assessed by means of AFLP technique using twenty sam- on the global 5-methylcytosine levels. First, the distribution ples from the C treatment. As stated before, this tool is of data sets was assessed in R by means of the based on the fragmentation of DNA using EcoRI and MseI Shapiro–Wilk test, and secondly homoscedasticity was restriction enzymes (Vos et al. 1995), which cut a specific analysed by using Bartlett or Levene test depending on sequence of nucleotides in the DNA. The statistical analysis whether data was normally distributed or not. For homo- was performed using the msap package but configured to scedastic data, differences between treatments were tested analyse AFLP data by using the logical value meth = by means of ANOVA (Analysis of Variance) or FALSE. As done for epigenetic analysis, genetic variations Kruskal–Wallis test, depending on whether data was nor- among plots were also explored by PCoA and the AMOVA. mally distributed or not, respectively, while for hetero- As mentioned above, genetic variation among treatments scedastic data a Welch’s ANOVA was used. Finally, post- was also assessed by using NML, whose banding patterns hoc analyses were performed when a dependent variable depend on variations at restriction sequence. This approach differed between climatic treatments, using Tukey’s, is appropriate when a high number of NML is detected in a Dunn–Bonferroni, and Games-Howell multiple compar- representative sample size (Pérez-Figueroa 2013; Watson isons test depending on the assessment used (ANOVA, et al. 2018). Kruskal–Wallis or Welch-ANOVA, respectively). The msap v1.1.9 package (Pérez-Figueroa 2013) devel- oped in R (R Core Team 2013) was used for the analysis of Results both MSAP and AFLP matrices. For MSAP, this package detects the activity of each restriction enzyme and classifies Plant phenotypic measurements each locus depending on the methylation state of the cytosines present in the target sequence (Supplementary Data Of the eight phenotypic traits considered in this study, four Table S2). Thus, the loci were classified as Methylation- exhibited significant differences among treatments (p value Susceptible Loci (MSL), which are used to assess < 0.05): Leaf Size (LS), Specific Leaf Area (SLA), Epigenetic responses of hare barley (Hordeum murinum subsp. leporinum) to climate change: an. . . 753

Fig. 1 Changes in some key phenotypic traits as a function of the letters above bars denote signiﬁcantly different groups after post-hoc climatic treatment. a Leaf size; b Speciﬁc Leaf Area; c Flowering test (FDR adj. p value < 0.05). onset and; d Seed mass. p p-value from overall effect test. Different

Flowering Onset and Seed Mass (Supplementary Data under control conditions. Finally, significant differences in Table S3). Among the three foliar traits, just LDMC did not seed mass were found among treatments, with WD plants significantly differ among treatments. Post-hoc comparisons producing seeds of higher weight, especially when com- indicated that plants subjected to drought (D and WD) had pared to plants subjected to drought (Fig. 1d). significantly higher LS and SLA than C plants (Figs. 1a, b). However, differences in LS between D and WD were not Intrapopulation genetic diversity significant at p value < 0.05, while differences in SLA between these treatments were statistically significant The AFLP technique was used for genotyping 20 indivi- (Supplementary Data Table S3). Plants exposed to warming duals from the C treatment. The number of loci produced by without drought (W) also tended to have higher SLA than C four selective primer combinations was 26, 26, 12 and 12 plants, but differences were not significant at p value 0.05. (for E34-M53, E35-M53, E34-M56 and E35-M56, respec- Regarding reproductive phenology, warming accelerated tively). The percentage of polymorphic bands from the total the onset of flowering in about ten days (both in W and WD of 75 loci analysed was 92%. Very low genetic divergence treatments; Fig. 1c) compared with those plants growing was found within this population, and differences among 754 V.íctor Chano et al.

Table 1 Analysis of molecular Source of variation d.f. SS MSS Variance Percentage of variation (%) Φ (p value) variance (AMOVA) for genetic ST (based on AFLP data) and AFLP analysis for genetic variability epigenetic (based on MSAP ns data) diversity. Among plots 5 60.96 12.19 0.3667 3.23 0.0323 (>0.05) Within plots 14 153.7 10.98 10.98 96.77 Total 19 214.6 11.3 MSAP analysis for epigenetic variability Among treatments 3 322.1 107.4 3.294 27.81 0.2781 (<0.0001) Within treatments 116 991.9 8.551 8.551 72.19 Total 119 1314 11.04 d.f. degrees of freedom, SS sum of squares, MSS mean sums of squares, ns not signiﬁcant.

Fig. 2 Principal Coordinate Analyses. a Results from PCoA to response to climatic treatments. Labels indicate the centroids, and analyse the genetic variation of hare barley among six experimental ellipses show the dispersion associated to each plot (a) or climatic plots. b Results from PCoA to analyse the epigenetic variation in treatment (b).

plots were not significant (ΦST value of 0.0323; p value = D (mean = 39.89%) and W (mean = 38.55%). Welch’s 0.278; Table 1). The AMOVA test revealed that genetic ANOVA (used for normally distributed heteroscedastic data) differences mostly occurred within plots (almost 99% of the resulted in a marginal significance (p value = 0.06), and variation), whereas just 1% was observed between plots. A post-hoc Games–Howell analysis for pair-treatments com- pairwise AMOVA also confirmed the lack of separation parisons revealed significant differences between WD plants between plots (Supplementary Table S4). In addition, the and those subjected to standalone stressors (p value < 0.1). PCoA did not show a clear separation of plants, and plots For the determination of epigenetic divergences among were overlapped with each other (Fig. 2a). The total var- treatments, a total of 354 loci in 120 individuals (42,480 iance explained reached 50.1%, with the first component of fragments) were analysed with the msap package using 12 the PCoA explaining 31.4%. primer combinations. From this total number of loci, 343 were classified as susceptible to be methylated (MSL), and Epigenetic variability 62 of them were polymorphic (18%). The Shannon diversity Index (S) for MSL was S = 0.51 ± 0.14 (mean ± SD). Climatic treatments induced significant differences among As shown in Table 1, differences in DNA methylation plants in both the percentage of cytosines methylated and the between treatments were highly significant, with a ΦST pattern of methylation. On the one hand, the quantification value from AMOVA for epigenetic variation among treat- of 5-mC by ELISA (Enzyme-Linked Immunosorbent Assay; ments of 0.2781 (p value < 0.0001). The remaining 11 loci Zymo) yielded percentages of methylated cytosines ranging were considered as NML, of which five were polymorphic from 30.8 to 76.4%. As shown in Fig. 3, the higher per- (45%). However, this approach is not appropriate for centages of methylated cytosines were found both in C inferring genetic variation when very low numbers of NML (mean = 52.03 %) and WD (mean = 55.08 %), compared to are detected. Epigenetic responses of hare barley (Hordeum murinum subsp. leporinum) to climate change: an. . . 755

Fig. 3 Global DNA methylation (mean of percentage ± SE) of methylated cytosines found for each climatic treatment. Different letters above bars denote signiﬁcantly different groups after post-hoc test (FDR adj. p value < 0.1).

Table 2 Results from pairwise analyses of molecular variance Table 3 Frequency (%) of methylation states at the target sequence for (AMOVA) between pairs of climatic treatments. each climatic treatment after multi-locus analysis. CDWWDDNA Control Drought Warming Warming + methylation state Drought C – D 0.2818* – Unmethylated 8.9 12.5 12.9 8.5 W 0.2856* 0.1882* – Hemimethylated 25.2 31.1 31.3 28.4 WD 0.2322* 0.3070* 0.3464* – Methylation of 42.7 40.7 41.6 37.0 internal cytosine Values correspond to Phi-statistic based on MSL loci. Uninformative 24.2 15.7 14.2 26.1 C control; D drought; W warming; WD warming + drought; *p value < 0.0001.

In Fig. 2b, a PCoA revealed multi-locus epigenetic dif- Plants subjected to different climatic treatments exhibited ferentiation between climatic treatments. Along the first distinct methylation patterns. Table 3 shows the frequency coordinate, which explained 19.7% of the total variance, a of the different methylation states detected in the different clear separation was observed between climatic treatments, climatic treatments. Differences among treatments were with W and D appearing in one side and C and WD in the small, although some remarkable differences were other side. Moreover, the C and WD treatments were observed: (i) the uninformative methylation state of MSAP separated along the second coordinate, explaining almost loci showed lower frequencies in D and W than in C and 9.5% of total variance. In addition, pairwise AMOVAs WD treatments; (ii) the unmethylated state in D and W was (Table 2)reflected the differentiation among treatments for higher than in C and WD; (iii) the hemimethylation state MSL diversity (epigenetic); epigenetic divergence between was more frequent in response to all climatic treatments treatments were significant for all pairwise comparisons. than in C, being higher in D and W than in WD; and 756 V.íctor Chano et al.

Fig. 4 Heatmap of 38 MSL obtained after a locus-by-locus Chi-squared test (p value < 0.001 after Benjamini and Hochberg adjustment), showing differences in methylation patterns between climatic treatments. Methylation states are f: full methylation; h: hemimethylated, i: inner cytosine methylation, and u: unmethylated. Samples (rows) and loci (columns) were clustered using the average linkage method.

(iv) methylation of inner cytosines was less frequent in after the onset of the experiment, the Geraniaceae Erodium response to the combined treatment WD. moschatum and especially Geranium dissectum decreased In addition, Fig. 4 shows a locus-by-locus Chi-squared in all treatments, particularly in those subjected to increased test analysis, which allowed to detect a group of 38 epiloci temperature and decreased rainfall. In contrast, hare barley, (11% of the total) with strong differentiation among climatic the grasses Avena barbata and Crepis capillaris, and the treatments (p value < 0.001 adjusted by Benjamini and Brassicaceae Sinapis alba tended to increase in 2018 Hochberg multitest), since the 120 plants analysed were (Fig. 5). Of the three grasses, only hare barley increased in broadly grouped in four clusters corresponding to each one abundance in all treatments. In fact, our study species of the climatic treatments (Fig. 4, row clustering). Within (together with S. alba) was the only dominant species that this sub-epiloci group, the proportions of DNA methylation increased its frequency in response to the combined effect states were similar to those shown in Table 3 for multi-locus of increased temperature and drought (i.e., WD treatment). analysis. Those epiloci showing uninformative state in C and WD plants, which can be also considered as full methylated (f), were mainly hemimethylated (h) in D and W plants. Discussion Figure 4 also shows that some unmethylated epiloci (u) in treated plants (i.e., plants subjected to D, W and WD treat- In this ﬁeld study we evaluated the phenotypic and mole- ments) were internally methylated (i) in C plants. Moreover, cular responses of hare barley to drier and warmer condi- some others unmethylated epiloci in C, D and W were found tions predicted by Climate Change models for the study internally methylated or hemymethylated in WD plants (i/h). area (Lindner et al. 2010). Our results show that hare barley displayed not only a remarkable phenotypic plasticity but Temporal changes in plant species frequencies also a high epigenetic diversity in response to increased temperature, decreased rainfall and the combination of both. The analysis of the variation in the frequency of hare barley Global methylation seems to be widely generalised in the relative to other co-dominant species indicated that 1 year genome of many plant species (Suzuki et al. 2008), being Epigenetic responses of hare barley (Hordeum murinum subsp. leporinum) to climate change: an. . . 757

Fig. 5 Changes in species frequency. Temporal variation in the relative frequency of the most dominant plant species when comparing plant communities in 2017 (ﬁrst spring after experimental treatment onset) and 2018 (second spring).

particularly frequent in species with high presence of Phenotypic responses to different climatic scenarios transposable elements such as hare barley (Jakob et al. 2004; Sharifi-Rigi et al. 2014), and large genomes such as Hare barley exhibited a large phenotypic plasticity in Zea mays (Diez et al. 2014). Specifically, our study species response to climatic conditions. Plants subjected to warmer was found to have the largest genome within the Hordeum conditions (in W and WD) flowered earlier, and those genus (Jakob et al. 2004). The epigenetic variation that subjected to combined stress (WD) showed a higher exhibited hare barley between treatments was higher than investment in leaf area per unit of leaf mass (i.e., higher the genetic diversity quantified for this species in the study SLA) and produced heavier seeds. Thus, the functional area, which is a common feature even in genetically diverse performance of hare barley was more sensitive to warming, plant species (Herrera and Bazaga 2010; Lira-Medeiros particularly in combination with reduced rainfall, than to et al. 2010; Richards et al. 2012). Even though genetic rainfall reduction alone. The increase in SLA with warming variation exists within plots, as suggested by the genetic is consistent with the effect of temperature in reducing leaf analysis of C plants, the overall low genetic diversity found thickness (Poorter et al. 2009; Lamaoui et al. 2018), between plots was an expected result taking into account the although this could also result from side effects of OTCs on low distances between plots and that the study species is air relative humidity affecting leaf area and cell wall characterised by anemochorous dispersal and cleistogamous thickness (Piikkia et al. 2008). Contrary to our initial pollination (Fotiou et al. 2007). However, uncertainty about expectation, rainfall reduction did not cancel out the posi- the potential effects of climatic treatments in selection tive effect of warming on SLA, and so plants subjected to remains, as germination and seedling survival of specific increased temperature and decreased rainfall did not pro- genotypes might be also influenced by different conditions duce leaves with trait values more associated to a water (Fernández-Pascual et al. 2013), so that the observed dif- conservation strategy. This increased biomass allocation to ferences could be the result of short-term, rapid adaptation, leaf area could potentially make plants more susceptible to instead of epigenetic modulation. Results from the present drought (Bongers et al. 2017). However, given the impor- study suggest that epigenetic diversity might be funda- tance of this attribute in whole-plant carbon gain and mental in acclimation to changing environmental condi- growth (e.g., Poorter et al. 2009), plants with high SLA tions, most likely by altering gene expression, which is in could be also more competitive in the uptake of resources, line with the results found in other studies (Bossdorf and not only light but also soil water and nutrients. The decline Zhang 2011; Li et al. 2012; Gayacharan 2013; Liu et al. in relative abundance of other co-occurring species such as 2018). Moreover, knowledge about phenotypic variation in G. dissectum and E. moschetum, particularly when exposed response to different climatic conditions can help to to potentially more stressful conditions, suggests that they implement management plans aimed at attenuating the were more affected by these climatic conditions than hare potential impact of climate change on Mediterranean barley. Therefore, hare barley could benefit from an dehesas, for example by adapting grazing management to increased availability of nutrients and water no longer used changes in plant growth and phenology induced by climate, by more stress-sensitive species in response to the isolated given the importance of these features on palatability and combined effects of warming and reduced rainfall. (Hussain and Durrani 2009). Similar indirect interactions between climatic changes and 758 V.íctor Chano et al. species performance have been discussed elsewhere (Ogaya drought in Arabidopsis thaliana also resulted in a unique et al. 2003; Seifan et al. 2010; Rodríguez-Calcerrada et al. response, with genes that were neither induced nor repres- 2013). In plant communities, species hierarchy and dom- sed by drought or heat stress alone responding to the inance patterns may vary depending on the species´ specific combination of both stresses (Rizhsky 2004). However, and competitive abilities under stress (e.g., Matías et al. 2018), despite epigenetics constitute an emerging discipline with with stress occasionally favouring non-dominant or sub- increasing interest in plant adaptation studies, few works ordinate species (Mariotte et al. 2013). have focused on the epigenetic control of gene expression in The potentially higher competitive ability detected for response to combined stresses. The comparison of com- hare barley under warmer and drier conditions might help bined stresses at the same time has been mostly analysed at us to explain the observed changes in its reproductive the post-transcriptomic level. For instance, Forestan et al. ecology in the WD treatment; plants subjected to the (2016), via RNA-seq analysis, identified long non-coding combined effects of decreased rainfall and increased tem- RNAs and small interfering RNA with specific roles in the perature advanced their flowering and produced bigger epigenetic regulation of gene expression in maize in seeds. The advanced flowering phenology with warming is response to drought and salt stress. To the best of our in agreement with the broadly known trend previously knowledge, the only study experimentally testing the reported by other studies (e.g., Whittington et al. 2015; combined effects of heat and drought at the epigenetic level Valencia et al. 2016; Moore and Lauenroth 2017). This was carried out by Liu et al. (2015), in which the enrich- early phenology likely allowed plants to produce bigger ment of Gene Ontology terms (i.e., terms representing gene seeds, since early flowering plants could potentially allocate products classified as cellular components, molecular resources to reproduction over an extended period of seed functions or biological processes) was related to epigenetic maturation (Wolkovich and Cleland 2014). The production regulation in a transcriptomic analysis conducted in Triti- of large seeds has been interpreted as a successful regen- cum aestivum. eration strategy under stressful conditions, where competi- In terms of methylation patterns, we also found strong tion among seedlings might arise due to high resource differences among plants as a function of the climatic limitation (Moles and Westoby 2004; Muller-Landau 2010). treatment. Plants subjected to decreased rainfall (i.e., D and WD treatments) exhibited the highest levels of the unme- Epigenetic responses to different climatic scenarios thylation state, while plants growing under C and W treatments showed the highest levels of the uninformative state. Both the level and pattern of methylation in hare barley As mentioned before, this methylation state is considered as varied substantially among the climatic treatments, with the uninformative due to the putative presence of changes in the combination of D and W inducing a clearly different epi- restriction targets. However, it might be also due to the full genetic response compared to that promoted by D and W methylation of the target, which corresponds to a repression when applied separately. In terms of global DNA methy- of the gene expression. Therefore, these results might imply lation levels, plants growing under control conditions dif- higher hypomethylation and subsequent upregulation of fered from those exposed to reduced rainfall (D) and putative responsive genes in those plants subjected to drier increased air temperature (W). These results were corro- conditions. This explanation is supported by previous stu- borated by pairwise AMOVA of epigenetic diversity among dies. For instance, drought-susceptible rice genotypes treatments, which allowed us to discern that the highest showed repression of drought-responsive genes via hyper- differences in epigenetic variation appeared between WD methylation under drought conditions, while drought- plants and those exposed to single D and W stresses, and tolerant genotypes exhibited hypomethylation and sub- then between C plant and those exposed to singles stresses sequent induction of gene expression (Gayacharan 2013). (W and D). Different studies have reported that the com- More recently, overall hypomethylation was detected in bination of biotic and/or abiotic stresses results in a unique drought-tolerant genotypes of wheat, in contrast to genome- response that is not the mere addition of the effects caused wide hypermethylated drought-susceptible genotypes (Kaur by each stress separately. In Kentucky bluegrass, for et al. 2018). Moreover, the locus-by-locus analysis per- example, the simultaneous presence of heat and drought formed in our study confirmed the observed frequencies of caused higher reductions in photosynthesis and leaf pho- methylation states for 41 epiloci significantly related to the tochemical efficiency than drought or heat alone (Jiang and four climatic treatments. Thus, higher proportion of unme- Huang 2000). In wheat, the synthesis of heat-shock proteins thylated loci were found for standalone treatments (D and induced by combined heat and drought was higher than W). In addition, the proportion of uninformative state after the application of only one of the stresses separately (putatively associated to full methylation and subsequent (Grigorova et al. 2011). Transcriptomic analysis by DNA gene repression) was higher in C plants when compared to chips revealed that acclimation to combined heat and treated plants (i.e., D, W and WD). These results seem to Epigenetic responses of hare barley (Hordeum murinum subsp. leporinum) to climate change: an. . . 759 confirm the existence of demethylation processes behind the climatic structures was possible thanks to DECAFUN project [grant overexpression of stress-responsive genes associated to number CGL2015-70123-R], funded by the Spanish Ministry of Sci- these 41 epiloci. ence and Universities.

Author contributions IMPR and MDHG installed the infrastructure of climate change simulation and performed phenotypic measurements. Conclusions VC conceived the original epigenetic research. VC, IMPR and JRC designed the final experiment. VC and TDF performed DNA extrac- The substantial differences in epigenetic diversity and the tions, AFLP, MSAP and ELISA experiments, and collected the genetic fi and epigenetic data. VC wrote the first version of the manuscript, and presence of speci c epigenetic patterns found in response to JRC and IMPR contributed substantially to the final version. All each climatic treatment suggest that molecular responses authors have read, edited and approved the final manuscript. underlying the reprogramming of gene expression and metabolic processes could be driving the functional stability Funding Open Access funding enabled and organized by of some dominant grass species (such as hare barley) under Projekt DEAL. different environmental scenarios (Zhu 2016). The potential role of epigenetic regulation as a mechanism of adaptation Compliance with ethical standards to new environmental conditions is reinforced by the low genetic diversity that exhibited the study species when Conflict of interest The authors declare no competing interest. comparing the four climatic treatments. Publisher’s note Springer Nature remains neutral with regard to These results will serve as foundation for further analysis jurisdictional claims in published maps and institutional affiliations. on how epigenetic variation underlies acclimation to different climatic conditions in hare barley. Understanding Open Access This article is licensed under a Creative Commons how genome-wide methylation affects plant function and Attribution 4.0 International License, which permits use, sharing, ultimately species frequencies will improve our predictions adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the on how climate change might alter plant community com- source, provide a link to the Creative Commons license, and indicate if position and ecosystem processes under future environ- changes were made. The images or other third party material in this mental scenarios. Transgenerational plasticity (e.g., via article are included in the article’s Creative Commons license, unless changes in seed size), influenced by epigenetic regulation indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended during adaptive responses of parental individuals to envir- use is not permitted by statutory regulation or exceeds the permitted onmental stress, opens new exploratory venues of research. use, you will need to obtain permission directly from the copyright Further analysis such as bisulfite sequencing and tran- holder. To view a copy of this license, visit http://creativecommons. scriptomic analysis (Metzger and Schulte 2017) will allow org/licenses/by/4.0/. fi correlating these epiloci to candidate genes with speci c References roles in phenotypic variation and functional responses to environmental changes. Alsdurf J, Anderson C, Siemens DH (2016) Epigenetics of drought- induced trans-generational plasticity: consequences for range Data availability limit development. Ann Bot 8:plv146 Anderson JT, Willis JH, Mitchell-Olds T (2011) Evolutionary genetics AFLP and MSAP fingerprints of hare barley control plants of plant adaptation. Trends Genet 27:258–66 Aragón-Gastélum JL, Flores J, Yáñez-Espinosa L, Badano E, and subjected to experimental conditions, respectively, can Ramírez-Tobías HM, Rodas-Ortíz JP et al. (2014) Induced be found at Zenodo repository under the following Digital climate change impairs photosynthetic performance in Echi- Object Identifier: 10.5281/zenodo.3531668. nocactus platyacanthus, an especially protected Mexican cactus species. Flora—Morphol Distrib Funct Ecol Plants – Acknowledgements We thank Dr. Carmen Collada (Universidad 209:499 503 Politécnica de Madrid) for her help and consent in using the UPM’s Banerjee A, Roychoudhury A (2017) Epigenetic regulation during fi facilities, Dr. María Ángeles Guevara (INIA) for her guidance in the salinity and drought stress in plants: histone modi cations and – design of lab protocols, and Dr. Andrés Pérez-Figueroa (CIIMAR) for DNA methylation. Plant Gene 11:199 204 his help with data analysis and guidance in the use of the msap Bartels A, Han Q, Nair P, Stacey L, Gaynier H, Mosley M et al. (2018) package. We also thank María Dolores Carbonero and Rafael Muñoz Dynamic DNA methylation in plant growth and development. Int for allowing us to install the infrastructures of temperature increase and J Mol Sci 19:2144 rainfall decrease in their dehesas. The authors would like to thank the Benjamini Y, Hochberg Y (1995) Controlling the False Discovery editor and two anonymous referees who kindly provided valuable Rate: a practical and powerful approach to multiple testing. J R – suggestions and comments to a previous version of the manuscript. Stat Soc 57:289 300 This work was supported by EPIGEHESA project [grant number Bonasio R, Tu S, Reinberg D (2010) Molecular signals of epigenetic – PRCV00434], funded by Fundación Biodiversidad, which is part of states. Science 330:612 6 the Spanish Ministry for Ecological Transition. The installation of Bongers FJ, Olmo M, Lopez-Iglesias B, Anten NPR, Villar R (2017) Drought responses, phenotypic plasticity and survival of 760 V.íctor Chano et al.

Mediterranean species in two different microclimatic sites. Plant perennial grass pastures of Western Oregon. J Chem Inform Biol 19:386–395 Model 33:1689–99 Bossdorf O, Richards CL, Pigliucci M (2008) Epigenetics for ecolo- Hussain F, Durrani MJ (2009) Seasonal availability, palatability and gists. Ecol Lett 11:106–115 animal preferences of forage plants in Harboi arid range land, Bossdorf O, Zhang Y (2011) A truly ecological epigenetics study. Mol Kalat, Pakistan. Pak J Bot 41:539–554 Ecol 20:1572–1574 Ibáñez I, Schupp EW (2001) Positive and negative interactions Chapin FS, Autumn K, Pugnaire F (1993) Evolution of suites of traits between environmental conditions affecting Cercocarpus ledifo- in response to environmental stress. Am Nat 142:78–92 lius seedling survival. Oecologia 129:543–550 Conrath U, Pieterse CMJ, Mauch-Mani B (2002) Priming in plant- Jakob SS, Meister A, Blattner FR (2004) The considerable genome size pathogen interactions. Trends Plant Sci 7:210–6 variation of Hordeum species (Poaceae) is linked to phylogeny, life Dabros A, Fyles JW (2010) Effects of open-top chambers and sub- form, ecology, and speciation rates. Mol Biol Evol 21:860–9 strate type on biogeochemical processes at disturbed boreal forest Jaskiewicz M, Conrath U, Peterhälnsel C (2011) Chromatin mod- sites in northwestern Quebec. Plant Soil 327:465–479 ification acts as a memory for systemic acquired resistance in the Delgado-Baquerizo M, Maestre FT, Rodríguez JGP, Gallardo A plant stress response. EMBO Rep 12:50–55 (2013) Biological soil crusts promote N accumulation in response Jeremias G, Barbosa J, Marques SM, Asselman J, Gonçalves FJM, to dew events in dryland soils. Soil Biol Biochem 62:22–27 Pereira JL (2018) Synthesizing the role of epigenetics in the Diez CM, Meca E, Tenaillon MI, Gaut BS (2014) Three groups of response and adaptation of species to climate change in fresh- transposable elements with contrasting copy number dynamics water ecosystems. Mol Ecol 27:2790–2806 and host responses in the maize (Zea mays ssp. mays) genome. Jiang Y, Huang B (2000) Effects of drought or heat stress alone and in PLoS Genet 10:e1004298 combination on Kentucky bluegrass. Crop Sci 40:1358–62 Doyle JJ, Doyle JL (1987) A rapid DNA isolation procedure for small Kaur A, Grewal A, Sharma P (2018) Comparative analysis of DNA quantities of fresh leaf tissue. Phytochem Bull 19:11–15 methylation changes in two contrasting wheat genotypes under Ewens WJ (2013) Genetic variation. In: Maloy S, Hughes K (eds) water deficit. Biol Plant 62:471–8 Brenner’s encyclopedia of genetics, pp 290–291 Kronholm I, Bassett A, Baulcombe D, Collins S (2017) Epigenetic and Excoffier L, Smouse PE, Quattro JM (1992) Analysis of molecular genetic contributions to adaptation in Chlamydomonas. Mol Biol variance inferred from metric distances among DNA haplotypes. Evol 34:2285–2306 Genetics 131:479–91 de la Riva EG, Tosto A, Pérez-Ramos IM, Navarro-Fernández CM, Olmo Fernández-Pascual E, Jiménez-Alfaro B, Caujapé-Castells J, Jaén- M, Anten NPR et al. (2016) A plant economics spectrum in Med- Molina R, Díaz TE (2013) A local dormancy cline is related to iterranean forests along environmental gradients: is there coordina- the seed maturation environment, population genetic composition tion among leaf, stem and root traits? J Veg Sci 27:187–199 and climate. Ann Bot 112:937–45 Lamaoui M, Jemo M, Datla R, Bekkaoui F (2018) Heat and drought Forestan C, Aiese Cigliano R, Farinati S, Lunardon A, Sanseverino W, stresses in crops and approaches for their mitigation. Front Chem Varotto S (2016) Stress-induced and epigenetic-mediated maize 6:26 transcriptome regulation study by means of transcriptome rean- Lampei C (2019) Multiple simultaneous treatments change plant notation and differential expression analysis. Sci Rep 6:30446 response from adaptive parental effects to within-generation Fotiou C, Damialis A, Krigas N, Vokou D (2007) Hordeum murinum plasticity, in Arabidopsis thaliana. Oikos 128:368–379 pollen as a contributor to pollinosis: important or trivial aero- Latzel V, Allan E, Bortolini Silveira A, Colot V, Fischer M, Bossdorf allergen? Allergy 62:180 O (2013) Epigenetic diversity increases the productivity and Freschet GT, Violle C, Bourget MY, Scherer-Lorenzen M, Fort F stability of plant populations. Nat Commun 4:2875 (2018) Allocation, morphology, physiology, architecture: the Laughlin DC, Leppert JJ, Moore MM, Sieg CH (2010) A multi-trait multiple facets of plant above- and below-ground responses to test of the leaf-height-seed plant strategy scheme with 133 species resource stress. N Phytol 219:1338–52 from a pine forest flora. Funct Ecol 24:493–501 Garnier E, Shipley B, Roumet C, Laurent G (2001) A standardized Li X, Zhu J, Hu F, Ge S, Ye M, Xiang H et al. (2012) Single-base protocol for the determination of specific leaf area and leaf dry resolution maps of cultivated and wild rice methylomes and matter content. Funct Ecol 15:688–95 regulatory roles of DNA methylation in plant gene expression. Gayacharan JA (2013) Epigenetic responses to drought stress in rice BMC Genom 2:300 (Oryza sativa L.). Physiol Mol Biol Plants 19:379–87 Lindner M, Maroschek M, Netherer S, Kremer A, Barbati A, Garcia- Gómez JM (2004) Importance of microhabitat and acorn burial on Gonzalo J et al. (2010) Climate change impacts, adaptive capa- Quercus ilex early recruitment: Non-additive effects on multiple city, and vulnerability of European forest ecosystems. Ecol demographic processes. Plant Ecol 172:287–297 Manag 259:698–709 Gower JC (1971) A general coefficient of similarity and some of its Lira-Medeiros CF, Parisod C, Fernandes RA, Mata CS, Cardoso MA, properties. Biometrics 27:857–71 Ferreira PCG (2010) Epigenetic variation in mangrove plants Grigorova B, Vaseva I, Demirevska K, Feller U (2011) Combined occurring in contrasting natural environment. PLoS One 5:e10326 drought and heat stress in wheat: changes in some heat shock Liu J, Feng L, Li J, He Z (2015) Genetic and epigenetic control of proteins. Biol Plant 55:105–111 plant heat responses. Front Plant Sci 6:267 Gu Z, Eils R, Schlesner M (2016) Complex heatmaps reveal patterns Liu G, Xia Y, Liu T, Dai S, Hou X (2018) The DNA methylome and and correlations in multidimensional genomic data. Bioinfor- association of differentially methylated regions with differential gene matics 32:2847–9 expression during heat stress in Brassica rapa. Int J Mol Sci 19:1414 Herrera CM, Bazaga P (2010) Epigenetic differentiation and rela- Liu Z, Xin M, Qin J, Peng H, Ni Z, Yao Y et al. (2015) Temporal tionship to adaptive genetic divergence in discrete populations of transcriptome profiling reveals expression partitioning of homo- the violet Viola cazorlensis. N Phytol 187:867–76 eologous genes contributing to heat and drought acclimation in Herrera CM, Pozo MI, Bazaga P (2012) Jack of all nectars, master of wheat (Triticum aestivum L.). BMC Plant Biol 15:1 most: DNA methylation and the epigenetic basis of niche width in Maestre FT, Escolar C, de Guevara ML, Quero JL, Lázaro R, a flower-living yeast. Mol Ecol 21:2602–16 Delgado-Baquerizo M et al. (2013) Changes in biocrust cover Hulting AG, Haavisto JL (2013) Hare barley (Hordeum murinum drive carbon cycle responses to climate change in drylands. Glob ssp. leporinum) biology and management in cool season Chang Biol 19:3835–3847 Epigenetic responses of hare barley (Hordeum murinum subsp. leporinum) to climate change: an. . . 761

Marion GM, Henry GHR, Freckman DW, Johnstone J, Jones G, Jones Piikkia K, De Temmerman L, Högy P, Pleijel H (2008) The open-top MH et al. (1997) Open-top designs for manipulating field tem- chamber impact on vapour pressure deficit and its consequences perature in high-latitude ecosystems. Glob Chang Biol 3:20–32 for stomatal ozone uptake. Atmos Environ 42:6513–22 Mariotte P, Vandenberghe C, Kardol P, Hagedorn F, Buttler A (2013) Poorter H, Niinemets Ü, Poorter L, Wright IJ, Villar R (2009) Causes Subordinate plant species enhance community resistance against and consequences of variation in leaf mass per area (LMA): A drought in semi‐natural grasslands (S Schwinning, Ed.). J Ecol meta-analysis. N Phytol 182:565–588 101:763–773 R Core Team (2013) R: a language and environment for statistical Mastan SG, Rathore MS, Bhatt VD, Yadav P, Chikara J (2012) computing. 55: 275–286. Assessment of changes in DNA methylation by methylation- Reyna-López GE, Simpson J, Ruiz-Herrera J (1997) Differences in sensitive amplification polymorphism in Jatropha curcas L. DNA methylation patterns are detectable during the dimorphic subjected to salinity stress. Gene 508:125–9 transition of fungi by amplification of restriction polymorphisms. Matías L, Godoy O, Gómez-Aparicio L, Pérez-Ramos IM (2018) An Mol Gen Genet 253:703–710 experimental extreme drought reduces the likelihood of species to Richards CL, Verhoeven KJF, Bossdorf O (2012) Evolutionary sig- coexist despite increasing intransitivity in competitive networks. J nificance of epigenetic variation. In: Plant Genome Diversity Ecol 106:826–837 Volume 1: Plant Genomes, their Residents, and their Evolu- Metzger DCH, Schulte PM (2017) Persistent and plastic effects of tionary Dynamics, pp 257–274 temperature on DNA methylation across the genome of three- Rizhsky L (2004) When defense pathways collide. the response of spine stickleback (Gasterosteus aculeatus). Proc R Soc B Biol Arabidopsis to a combination of drought and heat stress. Plant Sci 284:20171667 Physiol 134:1683–96 Mitchell P, Wardlaw T, Pinkard L (2015) Combined stresses in forests Rodríguez-Calcerrada J, Letts MG, Rolo V, Roset S, Rambal S (2013) (R Mahalingam, Ed.). Springer International Publishing, Multiyear impacts of partial throughfall exclusion on Buxus Switzerland sempervirens in a Mediterranean forest. Syst 22:202–213 Moles AT, Westoby M (2004) Seedling survival and seed size: a Seifan M, Tielbörger K, Kadmon R (2010) Direct and indirect inter- synthesis of the literature. J Ecol 92:372–383 actions among plants explain counterintuitive positive drought Moore LM, Lauenroth WK (2017) Differential effects of temperature effects on an eastern Mediterranean shrub species. Oikos and precipitation on early- vs. late-flowering species. Ecosphere 119:1601–9 8:e01819 Sharifi-Rigi P, Saeidi H, Rahiminejad MR (2014) Genetic diversity Muller-Landau HC (2010) The tolerance-fecundity trade-off and the and geographic distribution of variation of Hordeum murinum as maintenance of diversity in seed size. Proc Natl Acad Sci USA revealed by retroelement insertional polymorphisms in Iran. 107:4242–4247 Biology 69:469–77 Münzbergová Z, Latzel V, Šurinová M, Hadincová V (2019) DNA Shen X, De Jonge J, Forsberg SKG, Pettersson ME, Sheng Z, Hennig L methylation as a possible mechanism affecting ability of natural et al. (2014) Natural CMT2 variation is associated with genome- populations to adapt to changing climate. Oikos 128:124–34 wide methylation changes and temperature seasonality. PLoS Nicotra AB, Atkin OK, Bonser SP, Davidson AM, Finnegan EJ, Genet 10:e1004842 Mathesius U et al. (2010) Plant phenotypic plasticity in a chan- Suzuki MM, Bird A (2008) DNA methylation landscapes: Provocative ging climate. Trends Plant Sci 15:684–92 insights from epigenomics. Nat Rev Genet 9:465–76 Ogaya R, Peñuelas J, Martínez-Vilalta J, Mangirón M (2003) Effect of Tan MP (2010) Analysis of DNA methylation of maize in response to drought on diameter increment of Quercus ilex, Phillyrea latifo- osmotic and salt stress based on methylation-sensitive amplified lia, and Arbutus unedo in a holm oak forest of NE Spain. Ecol polymorphism. Plant Physiol Biochem 48:21–6 Manag 180:175–184 Tani E, Polidoros AN, Nianiou-Obeidat I, Tsaftaris AS (2005) DNA Olea L, San Miguel A (2006) The Spanish dehesa. A traditional methylation patterns are differently affected by planting density Mediterranean silvopastoral system linking production and nature in maize inbreeds and their hybrids. Maydica 50:19–23 conservation. In: Sustainable grassland productivity: Proceed- Valencia E, Méndez M, Saavedra N, Maestre FT (2016) Plant size and ings of the 21st General Meeting of the European Grassland leaf area influence phenological and reproductive responses to Federation warming in semiarid Mediterranean species. Perspect Plant Ecol Paun O, Bateman RM, Fay MF, Hedrén M, Civeyrel L, Chase MW Evol Syst 21:31–40 (2010) Stable epigenetic effects impact adaptation in allopoly- Verhoeven KJF, Jansen JJ, van Dijk PJ, Biere A (2010) Stress-induced ploid orchids (Dactylorhiza: Orchidaceae). Mol Biol Evol DNA methylation changes and their heritability in asexual dan- 27:2465–73 delions. N Phytol 185:1108–18 Pérez-Figueroa A (2013) msap: a tool for the statistical analysis of Vos P, Hogers R, Bleeker M, Reijans M, Van De Lee T, Hornes M methylation-sensitive amplified polymorphism data. Mol Ecol et al. (1995) AFLP: a new technique for DNA fingerprinting. Resour 13:522–527 Nucleic Acids Res 23:4407–14 Pérez-Ramos IM, Cambrollé J, Hidalgo-Galvez MD, Matías L, Wang WS, Pan YJ, Zhao XQ, Dwivedi D, Zhu LH, Ali J et al. (2011) Montero-Ramírez A, Santolaya S et al. (2020) Phenological rought-induced site-specific DNA methylation and its association responses to climate change in communities of plants species with drought tolerance in rice (Oryza sativa) L.). J Exp Bot with contrasting functional strategies. Environ Exp Bot 62:1951–60 170:103852 Watson RGA, Baldanzi S, Pérez-Figueroa A, Gouws G, Porri F (2018) Pérez-Ramos IM, Díaz-Delgado R, de la Riva EG, Villar R, Lloret F, Morphological and epigenetic variation in mussels from con- Marañon T (2017) Climate variability and community stability in trasting environments. Mar Biol 165:50 Mediterranean shrublands: the role of functional diversity and soil Westoby M (1998) A Leaf-Height-Seed (LHS) plant ecology strategy environment. J Ecol 105:1335–1346 scheme. Plant Soil 199:213–227 Pérez-Ramos IM, Matías L, Gómez-Aparicio L, Godoy Ó (2019) Whittington HR, Tilman D, Wragg PD, Powers JS, Browning DM Functional traits and phenotypic plasticity modulate species (2015) Phenological responses of prairie plants vary among coexistence across contrasting climatic conditions. Nat Commun species and year in a three-year experimental warming study. 10:2555 Ecosphere 6:1–15 762 V.íctor Chano et al.

Wolkovich EM, Cleland EE (2014) Phenological niches and the future LH (eds) Invasion genetics: the baker and Stebbins legacy, John of invaded ecosystems with climate change. AoB Plants 6:plu013 Wiley & Sons, Ltd, pp 328–340 Wright IJ, Reich PB, Westoby M, Ackerly DD, Baruch Z, Bongers F et al. Zhang X, Yazaki J, Sundaresan A, Cokus S, Chan SWL, Chen H et al. (2004) The worldwide leaf economics spectrum. Nature 428:821–7 (2006) Genome-wide high-resolution mapping and functional Zhang Y-Y, Parepa M, Fischer M, Bossdorf O (2016) Epigenetics of analysis of DNA methylation in arabidopsis. Cell 126:1189–201 colonizing species? A study of japanese knotweed in central Zhu JK (2016) Abiotic stress signaling and responses in plants. Cell Europe. In: Barrett SC., Colautti RI, Dlugosch KM, Rieseberg 167:313–24